Mental health apps frequently fail to outperform treatment controls — a predictable outcome based on decades of broader research that does not necessarily signal the apps are ineffective. Yet such controls remain the gold standard for establishing app effectiveness. This Comment argues that treatment controls often obscure the picture of whether apps work and calls for prioritization of alternative designs.
Second-generation antipsychotics (SGAs) are widely used in youths with severe mental illness but are frequently associated with cardiometabolic adverse effects. It remains unclear how much metabolic dysregulation results from body composition (BC) changes versus direct SGA effects. We followed 510 mostly antipsychotic-naive youths (male sex 55.7%, age of 4–17 years) initiating/reinitiating/switching to SGAs, with plasma concentration-verified adherence. BC and glucose/lipid metabolism changes were repeatedly assessed over 12 months. Metabolic changes were (1) controlled for and (2) correlated with body weight and fat mass, distinguishing BC-independent and BC-dependent metabolic dysregulation. BC-independent increases in total cholesterol, fasting glucose and triglycerides occurred especially with olanzapine and partially with risperidone but less with quetiapine and aripiprazole. Conversely, insulin resistance increased most in association with BC. SGAs dysregulate metabolism through distinct BC-independent and BC-dependent pathways, previously unexplored in large, well-characterized clinical samples. These findings highlight a need for cardiometabolic monitoring beyond weight in SGA-treated youths and for safer treatment alternatives. This study investigates cardiometabolic effects of second-generation antipsychotics in youth, distinguishing body composition-related changes from direct drug impacts. Findings reveal distinct metabolic dysregulation pathways, emphasizing the necessity for comprehensive monitoring and safer treatment options in clinical practice.
Psychiatry in the USA is at a crossroads: economic pressures have narrowed practice away from psychotherapy toward medication management while rapid growth in procedural treatments is reshaping the field. In this Comment, we argue that this shift requires a new identity, proposing interventional psychiatry that integrates psychiatric judgment with procedural care.
Substance use disorders (SUDs) remain widely undetected in healthcare settings, in part because comprehensive screening is hard to fit into brief clinical encounters. Machine learning offers a way to flag untreated SUDs from routinely collected data. Here we show that supervised machine learning models trained on electronic health record-like data from a nationally representative US survey can readily identify people with untreated alcohol use disorder, and exploratory models for rarer opioid and cocaine use disorders showed promise but were limited by small samples. Optimized for precision and recall, these models reduce the number of people who must be screened to find one untreated case relative to universal screening, in several cases by more than an order of magnitude, while remaining interpretable to show which features drive each prediction. Integrated into healthcare systems, such decision-support tools could improve early detection and treatment and reduce the substantial costs of untreated SUDs. In this study, the authors demonstrate that supervised machine learning models trained on electronic health record-like data can effectively identify untreated alcohol use disorder, potentially transforming early detection and treatment in healthcare systems.
Randomized controlled trials (RCTs) guide treatment decisions in youth with mental health disorders, yet concerns remain about their generalizability. Here we conducted a Preferred Reporting Items for Systematic Reviews and Meta-Analyses-compliant systematic review and meta-analysis (CRD42024629137) to assess the inclusivity and representativeness of RCTs on pharmacological and nutraceutical interventions in children and adolescents with Diagnostic and Statistical Manual of Mental Disorders and/or International Classification of Diseases-defined mental disorders. Random-effects meta-analyses were conducted by disorder, and temporal trends were assessed. Primary outcomes were the representativeness of RCTs with respect to sex (pooled proportion of female participants) and race/ethnicity (distribution of racial and ethnic groups); secondary outcomes included the exclusion of key clinical populations (for example, autistic individuals, those with intellectual disability and those with suicide risk). A total of 397 RCTs (46,989 participants; mean age 10.2 ± 3.1 years) were included, mostly targeting attention deficit hyperactivity disorder (34.5%) or autism spectrum disorder (29.7%). Female participants represented 28.3%. Only 53.7% of RCTs reported racial data; 75.0% were white, 10.1% Black, 6.2% Hispanic and 2.0% Asian. Autistic individuals and those with learning disabilities were excluded in 56.5% and 44.3% of trials, respectively; 76.0% of depression RCTs excluded participants at suicide risk. There is a persistent mismatch between the youth most affected by mental health disorders and those included in RCTs. Some vulnerable populations including people of colour, women and special populations seem to be under-represented, requiring more inclusive practice to support equitable and effective care. This systematic review and meta-analysis of 397 randomized controlled trials identifies an under-representation of women, people of colour and vulnerable youth populations in mental health research, highlighting the need for more inclusive methodologies in clinical trials.
Heterogeneity in clinical presentation and mechanisms of major depressive disorder (MDD) probably contributes to limited responses to current treatments in many patients. Identifying biologically meaningful phenotypes would constitute a major step towards the development of personalized treatment approaches. Brain-activity-based phenotyping offers a promising route toward this goal. In particular, brain oscillations—rhythmic patterns of neural activity that support information processing—have been implicated in depression but have not previously been used to define biological phenotypes of the disorder. Yet, no studies have used brain oscillations to identify biological depression phenotypes. Here we report data-driven identification of oscillation phenotypes for MDD. We conducted a cross-sectional study and collected resting-state magnetoencephalography (MEG), structural magnetic resonance imaging and clinical symptom data from 263 patients with MDD and 75 healthy controls. We assessed oscillation-based functional connectivity from source-reconstructed MEG data with two coupling-mode measures and computed their low-dimensional brain–symptom associations to obtain latent components. Using clustering methods on these components, we identified five depression phenotypes that were characterized by distinct spectral and spatial patterns and differentiated clinically unique symptom profiles. These findings suggest that MEG-based oscillatory connectivity captures clinically relevant heterogeneity in MDD and provides candidate mechanistic phenotypes for future validation and treatment-stratification studies. Using magnetoencephalography, structural magnetic resonance imaging and clinical symptom data from patients with major depressive disorder, the authors identify five depression phenotypes with distinct neural and symptom profiles.
Adults with attention deficit hyperactivity disorder (ADHD) have a higher prevalence and incidence of hypertension, but whether long-term cardiorenal trajectories after the initiation of antihypertensive medications differ by ADHD status is unclear. Here, in this nationwide retrospective cohort study using Dutch register data, we included 706,414 new users (52.2% female, aged 18–90 years) of antihypertensive medications without prior cardiovascular disease or chronic kidney disease (CKD); 10,689 had ADHD. Multistate modelling was used to characterize long-term cardiorenal illness trajectories and compare transition-specific rates between adults with and without ADHD. Adults with ADHD had higher adjusted rates of heart failure hospitalization (HHF; hazard ratio (HR), 1.46; 95% confidence interval (CI), 1.12–1.90) and stroke (HR, 1.19; 95% CI, 1.03–1.38) after antihypertensive medication initiation. They also had higher adjusted rates of cardiorenal death after HHF (HR, 1.79; 95% CI, 1.01–3.17) or CKD (HR, 2.57; 95% CI, 1.33–4.97) onset. Ten-year trajectories through HHF or CKD to cardiorenal death were more common in adults with ADHD. This research investigates long-term cardiorenal outcomes in adults initiating antihypertensive medications, revealing that those with ADHD experience significantly higher rates of heart failure hospitalization, stroke and cardiorenal death than counterparts without ADHD, using multistate modelling for analysis.
This Comment reviews pharmacists’ current contributions to psychiatric care in the USA and explains how their expertise, especially that of Board-Certified Psychiatric Pharmacists (BCPPs), positions them to have a key role in the safe, effective delivery of psychedelic therapies. Despite regulatory, reimbursement and training barriers, integrating BCPPs into this emerging field could improve patient safety, address workforce shortages and expand access to care.
In this Comment, we argue that nature and health research should move beyond theories focused on stress reduction and attention restoration to include capacity-building frameworks that emphasize flourishing, social connection and agency. We highlight nature-based social interventions as a means of strengthening individual, community and planetary well-being.
Adolescence represents a sensitive window for the maturation of executive function (EF), a core neurocognitive system underlying adaptive emotion regulation and social behavior. However, large-scale normative benchmarks of EF development and the extent to which deviations from these norms signal mental health vulnerability remain unclear. Here, using data from 33,622 Chinese adolescents aged 11–18 years, we mapped normative developmental trajectories of inhibitory control and working memory. Inhibitory control matured into late adolescence, whereas working memory plateaued earlier, accompanied by declining interindividual variability. Greater deviations below age-normed EF performance were associated with elevated peer and conduct problems, higher hyperactivity/inattention and reduced prosocial behavior, with these patterns being most pronounced during early adolescence and gradually attenuating thereafter. Replication in 11,549 US adolescents yielded convergent results. Together, these findings provide population-based benchmarks of EF development and highlight sensitive periods when neurocognitive deviations are most predictive of emerging psychopathology, informing identification and timing-specific prevention strategies. The authors model age-related changes in executive function during adolescence, focusing on inhibitory control and working memory in a large Chinese cohort with replication in a US sample.
Behavioral health reform in the USA is vulnerable to political change, limiting long-term benefits. In this Comment, we argue that durable behavioral health policy is a structural determinant of mental health and that embedding priorities in law and institutions is essential for continuity of care and lasting equity gains.
Internet-based psychological interventions can effectively reduce depressive symptoms in adults, but adherence remains challenging. In this preregistered individual participant data meta-analysis, we examined predictors of treatment adherence. We searched PubMed, Embase and PsycINFO on 6 February 2024 for randomized trials of internet-based interventions among adults with elevated depressive symptoms. We conducted a one-stage logit-link multilevel beta regression, with adherence defined as the proportion of completed modules post-intervention. This study included 71 trials (85 treatment arms, 8,082 participants). Lower adherence was associated with younger age (β = 0.005, standard error (SE) 0.002, P = 0.028), male gender (β = −0.163, SE 0.053, P = 0.002), lower education (β = −0.133, SE 0.05, P = 0.008) and employment (β = −0.113, SE 0.054, P = 0.037). No significant interactions were found between individual predictors and intervention format (guided versus self-guided). Higher adherence was associated with lower post-intervention depression severity, adjusting for baseline severity (β = −0.30, SE 0.04, P < 0.001). Identifying subgroups at risk of low adherence may inform targeted strategies to improve engagement and clinical effectiveness. Using data from 71 randomized controlled trials that involve 8,082 participants, the authors of this individual participant data meta-analysis examine individual- and study-level predictors of adherence to internet-based interventions for depression.
Higher socioeconomic status does not consistently protect against late-life depression. We propose that persistent pain may worsen depressive symptoms and amplify socioeconomic disparities through unequal access to pain management, reduced social engagement and limited financial resilience, offering a testable explanation beyond selective survival effects.
Depression can leave people feeling mentally ‘stuck’ in recurring negative thoughts, even when structured tests do not fully capture the problem. Using Munch’s Melancholy as a visual entry point, we argue that difficulty disengaging and working-memory discarding may sustain rumination, and that assessment should better capture these processes.
Predicting symptom change is a key goal of machine learning in mental health. However, models can seem more accurate than they are due to regression to the mean, a common statistical effect often overlooked in machine learning. Here we outline its implications and a simple framework for separating genuine prediction from statistical artefact.
Extreme weather and climate-related disasters are increasing in frequency, posing significant risks to mental health, particularly among people with pre-existing mental health conditions. However, little is known about the extent of these impacts or how social determinants shape vulnerability. Here, using population-based longitudinal data from Australia collected over a decade, we examined the mental health effects of climate-related disasters among individuals with and without pre-existing nervous, emotional or mental health conditions. Psychological distress and risk of moderate-to-severe mental disorders were assessed using the Kessler scale. Regression models accounting for fixed effects, confounding and clustering were applied. To estimate disaster impacts, exposed individuals were matched to comparable unexposed controls using one-to-five nearest-neighbor matching based on individual and area-level characteristics measured 1 year before the disaster. We also investigated whether modifiable social determinants—including residential stability, housing affordability, social support, economic safety nets and access to mental health services—altered disaster-related mental health outcomes. The results show that extreme weather and climate events were associated with higher psychological distress levels (β = 4.643, 95% confidence interval (CI) 0.891–8.395) and greater odds of moderate-to-severe mental disorders (odds ratio 2.233, 95% CI 1.000–4.984) among individuals with mental illness at the disaster year in comparison with a considerably smaller impact on those without. The psychological distress effects were significantly worsened when individuals with mental illness experienced residential instability (5.603, 95% CI 0.806–10.401), housing payment arrears (6.299, 95% CI 12.958–28.626), lower social support (4.775, 95% CI 0.972–8.577) or lack of mental health service contacts (6.640, 95% CI 0.576–12.703) during climate-related disasters. Findings underscore the heightened severity of mental health impacts among people with pre-existing mental illness and point to the need to integrate housing security, social support and sustained mental health services into climate disaster responses to reduce both the risk of mental disorders and climate-related inequities in this at-risk population. Using population-based longitudinal data from Australia, Li and Bentley evaluated how climate disasters affect mental health, showing that individuals with pre-existing conditions experience disproportionately worse outcomes, particularly under adverse social conditions such as housing instability and limited support.
Late-life depression is often viewed as a consequence of neurodegeneration or vascular brain injury, leading to therapeutic pessimism. We propose a recovery-oriented model that emphasizes brain plasticity and resilience, integrating biological, psychological and social factors to improve treatment and outcomes in older adults.
Suicide prevention research has largely focused on reducing despair and risk factors. Yet theory, evidence and clinical encounters show that suicidal people are driven not only by a will to die, but also by a will to live. This Comment urges greater attention to future life-focused psychological processes that support daily efforts to stay alive.
Interoception—the sensing and perception of internal viscera—is widely cast as a transdiagnostic mechanism linking brain–body interaction to mental illness. Prevailing models propose that altered interoceptive performance is associated with psychiatric vulnerability across a range of symptoms. Here we tested this hypothesis in a large cross-sectional community sample using psychophysically optimized tasks (N = 456 cardiac, N = 245 respiratory), with hierarchical Bayesian modeling and comprehensive symptom profiling. Contrary to this prediction, objective interoceptive performance metrics spanning sensitivity, precision and metacognition were largely unrelated to mental health dimensions across linear, categorical and network-based models. By contrast, self-reported interoceptive sensibility showed moderate associations with symptoms, but semantic similarity analyses suggest that these reflect higher-order interpretative or affective beliefs rather than interoceptive processing. These findings challenge prevailing views that objective interoceptive sensitivity is a broad marker of psychopathology, prompting reconsideration of how we measure and interpret interoception in mental health research. In this large-scale study, Banellis and colleagues investigate the links between mental health symptoms and interoception by measuring cardiac and respiratory interoceptive sensitivity and relating these to a multidimensional psychiatric profile.
